Tiny local decision models by Labqoat: calibrated answers in one forward pass, on NVIDIA, Mac and llama.cpp.
labqoat
company
AI & ML interests
None defined yet.
Recent Activity
Tiny local decision models by Labqoat: calibrated answers in one forward pass, on NVIDIA, Mac and llama.cpp.
A team of models on your device: qev-0.8b + a router + 2-bit LoRA experts (games, e2e). Mixture of Quantized LoRA Experts.
A 0.8B System One decision model: calibrated answers in one forward pass. Weights, MLX and GGUF builds, and our benchmarks.